ISSN: 2277-3754

ISO 9001:2008 Certified International Journal of Engineering and Innovative Technology (IJEIT)

Volume 3, Issue 10, April 2014

286

Abstract— This paper proposes an in-pipe inspection robot that

can adapt to changes in pipe diameter. The sophisticated design

and configuration of the robot enables it to travel autonomously

through pipes and adapt to diameters changing from 65mm to

220mm. It consists of two clamping modules and one module for

steering and locomotion. The clamping modules utilize flexible

spring sheets to enable the robot to dynamically adjust its radial

dimension and clamping force to attach to pipes of different inner

diameters and handle sudden or smooth changes in pipe

diameters. A middle module is engineered to enable the robot to

not only propel through the pipeline, but also to steer into the

desired direction when approaching elbows or junctions . The

robot finds its obstacle-free path using a set of sensors mounted on

the robot nose.

Index Terms—Autonomous robot, adaptation to pipe diameter,

pipeline inspection, steering.

I. INTRODUCTION

Pipelines are extensively used to carry natural gas and oil to

destinations throughout the world. They range from

high-pressure transmission lines to low-pressure distribution

lines. In order to prevent a failure/leakage in a pipeline due to

corrosion and/or high pressure, the interiors of the pipes need

to be routinely monitored and inspected to evaluate the need

to maintain or repair the pipeline, and to decide the most

effective means of doing so. Although cutting one or more

samples of the pipe is one of the most common methods used

for diagnosing a pipeline, its main disadvantage is attributed

to the high cost and downtime associated with the method,

other than the fact that this method does not provide a

comprehensive picture about the status of the whole pipe

because of the variations in corrosion rate or deposit buildup

found in different sections of a pipe. Moreover, many pipes

are located either underground or under water, or even in

places difficult to access. Therefore, many concepts of Pipe

Inspection Robots (PIR) have been utilized as alternative

methods [1-24] to access pipes from inside.

* This work was supported by Research Affairs at the

United Arab Emirates University under Research Grant #

1605-07-01-10 (PI: Basem F. Yousef). Features of the design

are reported and claimed in US provisional patent file No.

61577020. † Basem F. Yousef. Tel.: +971-50-1121 683; fax:

+971-3-762 3158. basem_yousef@uaeu.ac.ae.

Since space availability is one of the challenges that face

in-pipe robots[1], many concepts have been adopted to tackle

this challenge and resulted in developing different kinds of

in-pipe robots such as pig type , wheel type, caterpillar type,

wall-press type, walking type, inchworm type and screw

types, and the development of other robots that adopt

combinations of those concepts [1-24]. Moreover, pipe radii

vary by the usage and flow conditions of the pipes, hence

another main challenge that may pose technical difficulties to

PIRs is the ability to navigate inside a pipe of changing

diameter e.g. a pipe comprising segments of different

diameters. This challenge becomes more difficult when the

change in pipe diameter is 1) in step (not smooth) and; 2) the

diameter change is considerably large e.g., more than double

the smaller section. In these scenarios, when the robot passes

from a larger pipe to a smaller pipe diameter, it has to be able

to: a) find the hole location at the cross-section where the

pipes mate (e.g. up, down, right or left of the center of the

hosting pipe segment); b) steer into the hole direction; c)

change its radial dimension to negotiate the receiving pipe

diameter; and d) propel/move autonomously forward and

backward in a limited and constrained space.

Several concepts of PIRs have been developed to move

through a pipeline [2-7] with propulsion mechanisms that

depend on friction force which may not be always sufficient

to cause the motion. Other mechanisms were designed to

overcome those difficulties by utilizing sets of front and rear

rollers/wheels which are pushed against the interior wall of

the pipe and used to drive the robot [8-13]. Although some of

those PIRs could travel only in pipes of constant diameters

[8,9], others could handle only small changes in pipe

diameters since their wheels are pushed radially against the

pipe walls using compression springs of small strokes. Also

their springs usually exert smaller forces towards the end of

the spring stroke causing slipping of the wheels [10-15]. one

notable example on a PIR that can adapt to large changes in

pipe diameters is [13], which could handle a pipe change of

400-700 mm. It uses 3 pantograph mechanisms distributed

radially at 120 degrees to push the wheels against the pipe

walls, thus belongs to the wheel-driven robots. Its

configuration allows to maintain 3 contact points between the

wall and each of the 3 the drive mechanisms. However, the

contact points location pose constraints on the length of the

robot and its diameter, and it may cause collisions when

passing through pipe of large diameter to a smaller one

through a step connection.

Moreover, inch-worm robots [16-18] and snake-like robots

[19-21] utilize a serpentine motion to travel through

horizontal and curved pipes, but their designs allow for

Worm Robot with Dynamic Adaptation to Pipe

Diameter for In-Pipe Inspection1 1Basem F. Yousef

† and

2Nabil Bastaki

1Mechanical Engineering, United Arab Emirates University, basem_yousef@uaeu.ac.ac

2Electrical Engineering, United Arab Emirates University, nabil@uaeu.ac.ae

ISSN: 2277-3754

ISO 9001:2008 Certified International Journal of Engineering and Innovative Technology (IJEIT)

Volume 3, Issue 10, April 2014

287

limited radial and axial extension-contraction to provide

clamping force against pipe walls. Also, pipe crawlers

demonstrated good ability in dealing with the change in pipe

diameters [22, 23]. In spite of the advanced designs of those

robots, their intricate designs showed deficiencies in meeting

one or more of the intended tasks when navigating through a

pipeline especially when dealing with sudden changes in pipe

diameter.

In addition, magnetic wheels were utilized in [24] to

provide a wall climbing robot that may be used for inspecting

the interior of surfaces of gas-tanks in oversea ships, however

it lacks the ability to adapt to pipe diameter and it is material

dependent.

In this paper, we continue the effort of designing a

sophisticated engineering solution of a pipe inspection robot

that can efficiently adapt to changes of pipe diameter,

whether step or smooth, and to pass easily through elbows.

The initial conceptual design was reported in [25]. The novel

design of the proposed robot adopts an inchworm locomotion

concept to navigate autonomously through a pipeline.

Detailed description, performance analysis and testing is

explained in the subsequent sections.

II. ROBOT DESCRIPTION

The robot consists of three main segments, two identical

end clamps and a middle segment, in addition to a sensor

assembly to enable the path-realizing feature. The clamps

enable the robot to firmly attach to the inner walls of the pipe

exerting a preset constant clamping force in both straight pipe

segments and in elbows regardless of the diameter of the pipe

or elbow, while the middle segment provides means to a)

move and to b) steer the robot inside elbows or into the

direction of a clear path, as explained later. (Figs. 1 and 2).

The modules are connected by two ball joints to provide

flexibility needed to allow the robot to pass through turns. All

batteries and control drives are kept in a special compartment

connected to the middle segment.

Fig. 1. A 3D CAD model of the robot showing the different

modules

A. The Clamps

The main function of the clamps (Fig. 3a) is to provide

temporary contact with the inner walls of the pipe. Each

clamp assembly is built around a motor and consists of 6

flexible spring sheets connected at their ends to hexagonal

flanges by hinges. The flanges are connected with a power

screw and a DC motor which both are aligned to form the

clamp’s axis. While one hexagonal flange is fixed at the end

of the clamp, the other flange slides along the power screw

when the motor rotates. This movement causes the flexible

spring sheets to straighten or bend outwards to change the

radial dimension of the clamp, thus causing the sheets to

come in contact with the inner wall of the pipe. The closer the

flanges are, the wider the opening of the spring sheets is and

hence, the stronger the clamping force to the pipe will be

(Fig. 3b).

Controller and

power compartment

Steering and

drive module

End clamp

Front clamp

Sensor

assembly

Universal

joints

Micro DC

motors

Custom

ball joint

Fig. 2. Prototype of the PIR with inflated clamps for

attach-to-pipe pose.

ISSN: 2277-3754

ISO 9001:2008 Certified International Journal of Engineering and Innovative Technology (IJEIT)

Volume 3, Issue 10, April 2014

288

Fig. 3a. Components of a “Clamp”.

Fig. 3b. When slide flange moves from A to B, a wider radial

dimension for the clamp can be achieved.

However, since the robot is designed to move through

pipes of different diameters and/or, through a pipe of

changing diameter, this maximum clamping force can be

chosen to be constant regardless of the pipe diameter by

installing a mechanical torque limiter between the power

screw and the motor (Fig. 3a). When the torque limiter is

engaged, i.e., the torque on the shaft is less than a preset value

that depends on its specifications, the motor and the power

screw will be in direct drive. Once the motor torque exceeds

the limiter’s preset torque value, the torque limiter

disengages and disconnects the motor from the power screw

causing the motor to continue spinning but not the power

screw. Therefore, the hexagonal flange that slides along the

power screw will stop at a certain distance along the power

screw causing the spring sheets to apply a constant clamping

force to the pipe without the need to know the diameter of the

hosting pipe, and will set the clamping force at its maximum

value.

The controller, in turn, receives two encoder readings, one

from motor encoder 1 and the second is from encoder 2 that is

installed at the end of the power screw (Fig. 3a). When the

motor is engaged with the power screw, both encoders will

read the same angular velocity, but once disconnected, the

angular velocity of the power screw encoder will drop to

“zero” thus, the controller realizes that the clamp is tightened

with its maximum force to the pipe and ends the “clamping

task” and starts the “move task”.

Another feature of the clamp is the hole-finding sensor

assembly which is mounted on the frontal clamp (Figs. 2 and

4). The sensor assembly is designed to identify the existence

of obstacles and/or to realize the clear path during the robot’s

movement especially when the robot passes through an

elbow, or through local variations like a step change in pipe

diameter, eccentric sections, or connection into a network of

pipes through T-junctions.

Fig. 4. Ultrasound assembly set for hole- finding feature

The path-finding sensor assembly (Fig. 4) consists of three

parallel circular discs with an ultrasonic sensor placed on the

top disc, with the base plate attached to the nose of the frontal

clamp. Using a micro servo-motor, the top disc spins 270

degrees clockwise and counter-clockwise searching for a

clear path while the robot is in motion. The other two discs

are separated by 3 helical springs separated by 120 degrees to

keep the plates parallel, and each is concentric with a contact

sensor to detect any compression in the spring if it occurs. If

the robot fails to realize the clear path by the ultrasonic

sensors, signals from the contact sensors when the robot nose

hits an obstacle will guide the middle segment to steer the

robot to the appropriate path.

B. Middle Segment

This segment (Figs. 1 and 2) is responsible for two important

tasks, propulsion and steering. As shown in Fig. 5a, it

consists of three polymeric rods connected from one end to

universal joints which are fixed to an end flange, while at the

other end, the nylon rods connects to aluminum blocks that

can travel along motor-driven power screws. When the

blocks travel along the power screws, the nylon rods can

extend and retract, causing the middle segment to push/pull

its end flange along the robot’s length (Fig.5a). When the

driving micro-motors spin with the same speed, the nylon

rods extend or retract with the same linear rate which

pushes/pulls the end flange along a linear path. This enables

the robot, with the aid of the clamping segments to propel

along a linear path. On the other hand, if one of the rods

extends faster than the other two, the nylon rod bends causing

the end flange to steer (Fig. 5b). Hence, the middle module

enables the robot to propel and to steer to the desired

direction as instructed by the controller based on the direction

feedback obtained from the sensor assembly.

Fig. 5a. Middle segment’s details

ISSN: 2277-3754

ISO 9001:2008 Certified International Journal of Engineering and Innovative Technology (IJEIT)

Volume 3, Issue 10, April 2014

289

Fig. 5b. Robot can steer up to 90 degrees.

It is worth mentioning that the nylon rods are selected to be

strong enough to support the weight of the robot clamps and

provide sturdiness to the robot structure, but should be

flexible to allow bending when steering is needed.

Polyurethane rods were used since its properties meet such

requirements.

C. Motion Sequence and Pattern

The three modules of the robot (two end clamps and

middle segment) go through the following sequential

movement pattern that causes the robot to propel by an

inch-warm motion.

Initially, the front and rear clamps’ are in attach-to-pipe pose

where their spring sheets stick to the inner walls of the pipe,

while the middle module nylon rods are extended (Fig. 6a).

The motion starts by relaxing the rear clamp from the pipe

(Fig. 6b). This can be done by relaxing the spring sheets of

the clamp using the clamp’s power screw. Then the nylon

rods of the middle segment shrink/retract causing the rear

clamp to move forward (Fig. 6c). Once the rods are fully

extended, the rear clamp attaches to the pipe walls again (Fig.

6d) . Next, the front clamp releases the pipe (Fig. 6e) and the

nylon rods of the middle module extend, i.e. protrude,

pushing the front clamp forward (Fig. 6f). Once the nylon

rods are fully extended, the rear clamp is set to attach to the

pipe walls again (Fig. 6g).

Repeating that motion pattern can cause the robot to propel

in one direction while reversing the sequence can reverse the

motion direction.

III. MATH TESTS AND RESULTS

A. Motion in straight pipe segment

In order to determine the capability of moving along a

linear path, the robot was tested inside a 15 mm ID Plexiglas

pipe. The clamping force and propulsion speed depend on the

robot parameters and specifications of the components used

to build it as explained in this section.

Fig. 6. Motion sequence of the robot adopts inch-worm motion

pattern

IV. UNITS

However, in order for the robot to be able to travel vertically

through a pipe, the weight of the robot must be less than the

minimum of either the clamping force of a single clamp or the

drag force generated by the middle segment, whichever is

less. Therefore, those forces were measured as follows:

To measure the clamping force, the clamps were attached to

the pipe with maximum force which is dependent on the

torque limiter’s capacity as mentioned earlier, then the robot

nose was hooked to a load cell and pulled by a test motor by

gradually increasing the motor torque while monitoring the

position of the robot inside the pipe. The clamping force was

measured by the load cell as the force sufficient to slide the

robot. The experimental setup used to conduct the test is

shown in (Fig. 7a) and explained further in the illustrative

diagram of Fig. 7b.

Also the same setup was used to measure the drag force.

For this, the test motor brake was engaged and one of the

clamps is set free while the other is clamped to the pipe, and

the drag force was measured by the load cell as the maximum

pulling force caused by the middle segment without slippage

of the attached clamp. (Fig. 7c). Test results showed that, for

its current configuration, the robot clamps together can

provide a 76 N, i.e., 38 N/clamp, while the drag force was

found to be 83N. It is obvious that the critical force to

compare with the robot’s payload in the case of climbing a

vertical pipe is the clamping force. Since the weight of the

robot, 3.55 kg = 34.8 N is less than the clamping force, the

robot could travel vertically through the pipe. Note that the

roughness of the inner surface of the pipe will affect these

values because it affects both the clamping and the drag

(a)

(b)

(c)

(d)

(e)

(f)

(g) = (a)

ISSN: 2277-3754

ISO 9001:2008 Certified International Journal of Engineering and Innovative Technology (IJEIT)

Volume 3, Issue 10, April 2014

290

forces. Also note that the drag and clamping forces that

suffice travelling through a vertical pipe will be adequate to

enable travelling through a horizontal pipe.

Fig. 7. (a) Test bed used for clamping force measurement (b)

Clamping force measurement, (c) Drag force measurement.

Moreover, the clamping force to the walls can be adjusted or

changed by changing one or more of the following

parameters/specifications: 1) the spring sheets’ thickness,

width and/or material; 2) the torque limiter’s set-torque. The

robot’s current configuration uses 0.63 mm thick x 10 mm

wide stainless steel spring sheets, and the selected torque

limiter’s capacity is 392 mN･m. Furthermore, the length of

the spring sheets controls the robot’s capability to handle

larger pipe diameters. Fig. 8 shows the increase in the contact

surface between the spring sheets of the clamp and the pipe

surface as a result of using a torque limiter of higher capacity.

Fig. 8. Flat spring sheet with larger surface of contact with the

pipe results from using a torque limiter of larger capacity.

The propulsion speed was calculated based on the total

time required to complete the full cycle of an inch-worm

motion pattern as described in the previous section and found

to be 2 mm/s. However, the speed can be improved by

minimizing the controller’s wasted time. For instance,

attaching/releasing the clamps to the pipe can be performed

by spinning the clamp motor few spins sufficient to apply

release the necessary force or to release the clamp instead of

completely flattening the spring sheets. Also, this speed can

be improved further by using micro-motors with higher RPM

to drive the elastic rods of the middle segment and/or by

selecting power screws with larger pitch or lead. Also power

transmission gears will be considered in future versions.

Table I summarizes the specifications of the robot based on

its current configuration.

Table I. Specifications of the Robot.

Requirement considered Specification

Weight < 3.6 kg

Material Brass and aluminum

Architecture Modular: Clamp- steering/propulsion

segment- Clamp

Motion type Inch-worm

No. of Actuators 5

Power supply Self-powered , 8-AA batteries+ 2X 9V

batteries

Dimensions: (Length)

(Outer diameter)

109 cm (fully retracted)

117.5 cm (fully extended)

(min) 6.5 cm- fully collapsed clamps (max) 22 cm- fully expanded clamps

Max linear stroke 8.5 cm

Linear propulsion speed

(tested in 15 cm ID pipe)

2 mm/s

Clamping force (tested in a

15 cm ID Plexiglas pipe )

76 N

Max turning angle 90 degrees

Operating mode Autonomous

B. Motion in turns

The motion pattern can be briefed as follows: the middle

module is responsible for pushing the front of the robot, to

steer it, and to drag the rear section forward by active

actuators. However, the rear section comprises two passive

ball joints (Figs. 1 and 2) to allow the robot body to bend

when passing through turns. Fig. 9 shows the robot

immerging from a straight pipe to pass through an elbow

during experimental test of the capability of the robot to

handle turns. For this experiment, a pipe/elbow diameter of

Dp=180 mm and an elbow radius re= 270 mm where used

(in urban gas pipelines, is usually 1.5 ).

Pipeline configurations and dimensions impose

geometrical constraints on the robot dimensions. S Roh and

H Choi provided detailed analysis on such restrictions [12].

The analysis for the proposed robot provided in this section is

based on their approach.

Since the robot comprises a series of rigid bodies, the

analysis will consider the longest rigid part which, if passes

the elbow, all other parts can pass successfully. Referring to

Fig. 2, our target part is the frontal clamp that includes the

(a)

(b)

(c)

ISSN: 2277-3754

ISO 9001:2008 Certified International Journal of Engineering and Innovative Technology (IJEIT)

Volume 3, Issue 10, April 2014

291

sensor assembly with a total length of 340 mm. The worst

scenario is when the part makes 45 with the tangent to the

elbow as shown in Fig. 10.

Fig. 10. Collapsed Clamp inside an elbow. Dr,min= 65mm.

From the geometry, the length of the robot’s longest rigid part

is given as follows: in case the length exceeds the distance

, the robot’s min. diameter has the range of

[√

(

) (

)] (1)

where is the radius of the elbow from center to its middle,

and is the pipe diameter.

The length of the robot is given by

√ [

√

(

)] (2)

setting , then

(3)

Considering this range of from (3) in (1)

√ (4)

The second case is when the length of the longest rigid body

is within the elbow curvature, i.e. the length is less than the

distance , then

[√

(

) (

)] (5)

and √[ (

)

(

)

] (6)

again, when setting and considering the range of

given in (5),

√ (7)

√ (8)

The preceding analysis assumes that the robot’s radial

dimension is slightly smaller than the height h. From

the equations, it is obvious that if satisfies (8), then the

robot can pass through straight segments and elbows,

provided that the robot’s configuration can handle the pipe

diameter.

Therefore, the maximum pipe diameter depends on the 1)

radial dimension that the robot’s configuration can handle,

and; 2) that the robot’s and satisfy the constraints

imposed by (7) and (8). However, (8) confirms that the range

of that can pass an elbow of =180 mm, which was used

for the test elbow, is , and this range includes

=340 mm (that is the longest rigid body in our robot). Note

that the robot’s configuration can handle pipe diameters

between 65-220 mm, which includes the 180 mm pipe/elbow

diameter.

One of the main advantages of the proposed robot would

be obvious when passing through bends as the flexible spring

sheets can take the shape of the hosting pipe when clamping

to the pipe.

Fig. 9. Robot emerging from a straight pipe then passing through an elbow, Dp=180 mm, re=1.5Dp= 270 mm

ISSN: 2277-3754

ISO 9001:2008 Certified International Journal of Engineering and Innovative Technology (IJEIT)

Volume 3, Issue 10, April 2014

292

V. CONCLUSION

A novel in-pipe inspection robot has been designed and

constructed. The robot’s features enable it to navigate

autonomously through a pipeline. The robot that is

configured with three main modules, two clamps and one

drive segment adopts inch-worm motion pattern. Its

clamping modules enable the robot to dynamically adjust its

radial dimension between 65-220 mm, and clamping force to

attach to pipes of different inner diameters and overcome the

obstacles associated with changes in pipe diameters, thus it

can travel through pipelines that comprise segments of

varying inner diameters. It comprises a sophisticated middle

module that enables it to propel, and to steer into the desired

direction when approaching an elbow or a junction. In

addition, the robot is equipped with a hole-finding sensor

assembly that allows it to realize the clear path along turns

and T-junctions.

REFERENCES

[1] N. Saga, T. Nakamura, “Development of a peristaltic crawling

robot using magnetic fluid on the basis of the locomotion

mechanism of the earthworm”, Smart Materials and Structures,

2004,13, pp.566–569.

[2] F. Kirchner and J. Hertzberg “A Prototype Study of an

Autonomous Robot Platform for Sewerage System”,

Maintenance Autonomous Robots, 1997; 4(4), p. 319-331.

[3] HB. Kuntze and H. Haffner, “Experiences with the development

of a robot for smart multisensoric pipe inspection”, Proc. IEEE

International Conf. on Robotic and Automation, 1998. p.

1773-8.

[4] SG Roh and HR Choi, “Differential-drive in-pipe robot for

moving inside urban gas pipelines”, IEEE Transactions on

Robotics, 2005; 21(1), pp. 1-17.

[5] M. Muramatsu, N. Namiki and R. Koyama, “Autonomous

mobile robot in pipe for piping operations”, Proc. IEEE/RSJ

International Conf. on Intelligent Robots and Systems, 2000.

pp. 2166-71.

[6] M. Moghaddam and M. R. Tafti, "Design, modeling and

prototyping of a pipe inspection robot", 22nd Int. Symp. on

Automation and Robotics in Construction (ISARC), 2005.

[7] Yunwei Zhang, and Guozheng Yan, "In-pipe inspection robot

with active pipe-diameter adaptability and automatic tractive

force adjusting",Mechanism and Machine Theory, 2007, 42,

pp. 1618-1631.

[8] HR. Choi and SM. Ryew “Robotic system with active steering

capability for internal inspection of urban gas pipelines”,

Mechatronics, 2002; 16(12), pp. 713-36.

[9] Z. Zhu and Z. Pan, “Miniature pipe robots”, Industrial Robot:

An International Journal, 2003; 30(6), pp. 575–83.

[10] I. Hayashi, N. Iwatsuki, K. Morikawa and M. Ogata, “An

in-pipe operation microrobot based on the principle of screw”,

International Symposium on Micromechatronics and Human

Science, 1997, pp. 125-9.

[11] M. Horodinca, I. Dorftei, E. Mignon and A. Preumont, “A

simple architecture for in-pipe inspection robots”, Proc.

International Colloquium on Mobile and Autonomous

Systems, 2002, pp. 61-4.

[12] S. Roh and H. R. Choi, "Differential-drive in-pipe robot for

moving inside urban gas pipelines", IEEE Transactions on

Robotics, 2005, 21(1), pp 1-17.

[13] J. Park, T. Kim, H. Yang, “Development of an actively

adaptable in-pipe robot”, Proc. IEEE International Conference

on Mechatronics (ICM), 2009.

[14] K. Suzumori, T. Miyagawa, M. Kimura, Y. Hasegawa, “Micro

Inspection Robot for 1-in Pipes”,IEEE/ASME Transactions on

Mechatronics, 4(3), 1999, pp.286-292.

[15] S. Hirose, H. Ohno, T. Mitsui and K. Suyama, “Design of

In-Pipe Inspection Vehicles for ø25, ø 50, ø150 Pipes”, Journal

of Robotics and Mechatronics, 12(3), 2000, pp. 310-317.

[16] T. Fukuda, H. Hosokai, M. Uemura, “Rubber gas actuator

driven by hydrogen storage alloy for in-pipe inspection mobile

robot with flexible structure”, Proc. IEEE International Conf.

on Robotics and Automation, 1989, p. 1847–52.

[17] S. Aoshima, T. Tsujimura and Yabuta “A miniature mobile

robot using piezo vibration for mobility in a thin tube”,

Transactions of ASME, Journal of Dynamic Systems,

Measurements and Control, 1993, 115, p. 270-8.

[18] C. Anthierens, A. Ciftci and M. Betemps, “Design of an

electro pneumatic micro robot for in-pipe inspection”, Proc.

IEEE International Symposium on Industrial Electronics, 2,

1999, pp. 968-72.

[19] KU. Scholl, V. Kepplin, K. Berns and R. Dillmann,

“Controlling a multi-joint robot for autonomous sewer

inspection”, Proc. IEEE International Conf. on Robotics and

Automation, 2000, 2, p. 1701-6.

[20] S. Wakimoto, J. Nakajima and M. Takata, “ A micro snake-like

robot for small pipe inspection. Proc. International Symposium

on Micro Mechatronics and Human Science, 2003, pp. 303-8.

[21] J. Borenstein, G. Granosik and M. Hansen, “The OmniTread

serpentine robot for industrial inspection and surveillance”,

International Journal on Industrial Robots, Special Issue on

Mobile Robots, 2005, 32(2), pp.139-48.

[22] A. Zagler and F. Pfeiffer, “MORITZ” a pipe crawler for tube

junctions”. Proc. IEEE International Conf. on Robotics and

Automation, 2003, pp. 2954-60.

[23] W. Neubauer, “A spider-like robot that climbs vertically in

ducts or pipes”, Proc. IEEE/RSJ International Conference on

Intelligent Robots and Systems, 1994, pp. 1178-85.

[24] W. Fischer, F. Tâche, and R. Siegwart, “Inspection system for

very thin and fragile surfaces, based on a pair of wall climbing

robots with magnetic wheels,” in Proc. IEEE/RSJ Int. Conf.

Intelligent Robots and Systems, San Diego, CA, USA, 2007,

pp. 1216-1221.

[25] Basem F. Yousef, et. al., “IN-PIPE INSPECTION ROBOT”.

In Proc. of the 15th Int. Conf. on Climbing and Walking Robots

and the Support Technologies for Mobile Machines, BA, USA,

2012, pp. 289-296.